90 Northern Andean - Search and Research

Mike Adams Photograph 3/

The high cloud-forests and moors of the northern Andes are a paradise for the biologist. Here in Colombia and Venezuela, where the great Andean chain fragments, each mountain range is like an island and each island has its own ofanimals and plants. Colombia has five major mountain islands, three of them well over 5000 m high, separated from each other by deep valleys and low, narrow ridges; Venezuela has one 5000m chain emerging from Colombia's Eastern Cordillera. George Bernard (now a photographer) and I began our explorations of these mountains in 1971 and I have been there now six times. "Vith rucksacks, tent, nets and envelopes stored in boxes, we have scoured all six of the north Andean 'islands' for their high-altitude butterflies. Early on we discO\'ered several important things. firstly, as you climb the mountain-sides, the numbers of butterfly species diminish, but the proportion of endemic species increases dramatically. Secondly, virtually all the high-altitude butterflies require untouched cloud-forest or unculti\-ated moorland, and these - especially the former - are hard to find and hard to get to. Thirdly, ob\'iously few collectOrs haye bothered with the high ele\'ations, because cool rain and mist seem incompatible with butterfly life and most ofthe species up there are browns: so down they ha\'e gone, back to the mosquitoes and sandflies and the bright and beautiful butterflies of the lowlands, where the sun shines and your rucksack sticks to your back. Fresh (i'om university and Alpine butterflies, we chose to start with the Sierra ~e\'ada de Santa Marta, an isolated montane island ifever there was one - the area ofVorkshire, 5775 m high, rising plumb out ofthe Caribbean and surrounded by plains a few hundred metres above sea-level at the most. We bought a donkey - 'Balthazar Burro' - for our first attempt at the cloud-forests. For £10 we thought we had a bargain, but the efkminate, worm-ridden beast was a liability from the start, and on the first evening we were already carrying the rucksacks ourselves. Without a map and well and truly lost on the second day, we entered the territory of the Ica (or Arhuaco) Indians, tall, proud people wearing domed flaxen hats, woollen ponchos and coloured bags, chewing coca leaves and wary of strangers. The burro fell on' the path near Donachui and ended up on its hind legs on a ledge 2 m vertically below; in the hot sun - we were still at only about 1400m - we made a diagonal cut with machetes to guide the back. At Sogrome (1700m) we could see the cloud-forests above, but, the Indians told us, the route to Tromba was very dangerous and would take us three days; we were there in two hours' In idyllic Tromba, camped on the softest ofgrass beside a derelict lea hut on the banks of a sparkling torrent (2300m), we had our first taste of cloud-forest ~ORTHER~ ANDEAC'I BCTfERFLlES 91 browns. No books exist on Colombian butterflies, so we gave them our own nicknames and made our own judgements about what was interesting; we guessed that one ofour browns was rare, but little did we then know that it was not only a species new to science, but a new genus as well (Arhuaco ica Adams & Bernard). Getting to the uppermost forests on the western flank ofthe Santa Marta range was harder and we needed three attempts, foiled by losing our way, fever and lack ofwater (yes, this is possible in cloud-forest when the path relentlessly follows a knife-edged ridge and there is no-one around to tell you where the sources are!). But we finally made it, in two days from San Pedro de la Sierra, in 1972. Our prize up there, in the forest right at the tree-line, was another new species, Pedaliodes cebolLeta (meaning 'little onion' and named after the farmstead down below). Further up, in the grassy moors called the 'paramo', we found two species of browns: one of them was white, belonging to the genus Lymanopoda that in later years became almost an obsession - a biologist interested, as I am, in geographical distributions of related species and in their evolutionary histories could not ask for a more perfect group oforganisms to give him ideas and illustrate his theories. Our troubles in the Santa Marta massifwere by no means over. Trying to reach the paramo from the south-east, via Maruamaque, we were 'sent to Coventry' by dozens of Kogui Indians massing for a meeting in their religious centre, Chenducua, and were advised by a North American missionary to retrace our steps rapidly: top oftheir agenda was how to deal more forcefully with intruders bent on stealing their tribal heirlooms. Frankly, they would not believe that we were purely lepidopterists. The lea Indians were also flexing their muscles in 1972, determined to show that they should choose who enters their land, not the authorities down in non-Indian Valledupar; so, when we reached Donachui armed with all the proper documentation, the whole \'illage assembled in a long hut, men down one side, women d.own the other and we in the middle. facing the chiefshaman and a self-appointed political leader. Long speeches were delivered in pidgin Spanish, giving, we supposed. the reasons why we were not welcome and should heaq back down the valley (unless we paid some £200). A month later, we crept through Donachui long after dark, guided by a friendly Ica and his sturdy mule and stayed the night in one ofhis family huts at Sogrome along with his father, the local shaman. Once more we passed through Tromba and trudged up to the 4000m pass above it, camping at 3950m with Adalberto wedged in the tent between us. The identification ofour material (over 600 species from 6 months' collecting in the Sierra Nevada de Santa Marta) had to be attempted by ourselves, using the collections and literature at the British Museum ofNatural History (which, since 1975, has purchased all our butterflies from Colombia and Venezuela). It was then that we realised we had numerous new species - far too many to work on ourselves - and I began to concentrate on the browns, embarking on the first of our papers on the 'Pronophilini' ofthe northern Andes.' The intricacies of are a far cry from the joys of searching for and catching the butterflies one is describing: The evolutionary relationships of the 13 endemic species are analysed. Three new genera, three new species, and four new subspecies are described. Four new synonymies are established, the status of 92 TH E .-\LPI:"EJOL'R:"..1,L three taxa revised, six new combinations made, and one original combination re-instated (from the abstract at the start ofour first paper). The white Lymarwpoda, for instance, had been named L. nevada in 1924. In 1927, another German named the same thing L. kruegeri, obviously unaware that Kruger himself had already described it. Then in 1931, Schultze called it Sabatoga nevadmsis - yet another name for the same species and put into a different genus to boot' Sortingout valid names - the first one always stands- is essential before you can begin to discuss the species' relationships in print and has taken up a lot ofmy non-teaching time since 1972. And if a species is new and undescribed, you cannot just arbitrarily give it a name and refer to it as such in your next paper: it has to be formally described. 'Forewing length: 33-35mm. Upperside unicolorous dark brown, with postdiscal black spots showing through from underside, on the forewing in cells Cu2 to Ml, and on the hindwing in cells Cu2 to R5 .. .' (Arhuaco ica). A halotype has to be selected and designated ('Holotype male, COLOMBIA: W. Sierra Nevada de Santa Marta, E. ofSan Pedro, 2700m, 6.viii.72 (M.J.A.),), and the full data given for all other specimens caught (the paratypes). The species' relationships need to have been established, its closest relatives determined by a variety ofcriteria, including the shapes of their males' genitalia. This is the fascinating but dry world of the museum man. But back now to the field' From the Santa Marta range, you can see the northernmost finger ofthe Andes proper - the Serrania de Valledupar, the highest part of the Sierra de Perija. Straddling the sensitive border with Venezuela, most of its few high-altitude human inhabitants are smugglers and renegades, and there are also persistent rumours about the savagery of the local Motilone Indians. No wonder no collector had ever been up there before! We reached its cloud-forest on a moonlit Christmas Eve, 1974, ready to sink into exhausted, thirsty sleep on the path when we spotted an abandoned homestead in a forest clearing. Our blistered feet prevented us from ascending above the lower cloud-forests that time, but we returned successfully a month later, using a 20-year-old mule ofgreat depend­ ability. On the Colombian side, above Manaure, the tree-line is low, down at 2800m, and the forest - bristling with mammals such as howler and spider monkeys, bears and a fierce blue carnivorous 'fox' - gives way to the 'Sabana Rubia', a gently-sloping, marshy expanse of 'pseudo-paramo', leading up to the border. Here, and in the lichen-encrusted uppermost forests at 3050m on the Venezuelan side, we found most of the 14 unknown species of browns that came our way in this amazing mountain range.l Condors circled overhead. Next on our itinerary was the Venezuelan Andean chain, the Cordillera de Merida. South of the town of Merida, the longest 'teleferico' in the world takes you in four stages to 4765m (when it works). North of the town, the cloud-forest and paramo are much more virgin, but you have to walk there from 1600m. There's no white Lymanopoda in this range, but the paramo harbours three very extraordinary brown species. Two of them illustrate a common feature of the paramo fauna, the existence of'relicts', species with no known relatives on earth which may have moved from the south into the new grasslands that formed above the forests as the Andes rose. Diaphanos huberi, discovered in 1971 by Professor Huber and collected by ourselves in 1975 and 1977 between 3500 and 4000 m, is 31 nit sa 111 ius-group Dj I.ymanopoda bUllerfly species in Colombia. The Ill'O colnll1nsr!f -I lift species lil'ing ill atijartlll bantis ojalliluti, up Iheir mounlain slopes. S/Jtcies rdarrti 10 inlht Itxlart ne"ada (lop Idl). paralllera (lop righl). pieridina (bol/om Idl). samius (bOl/om righl) alltilaClea (lI1ititil, ".~Itt). I'h",,, .I/,A, .Id"m' "iORTHER'" 1\1\0E.-\:\ BU1TERFLI ES 93 uniquely tiny and almost transparent and flies like a micro-moth.:l Redonda empetrus is large and pale brown, with broad rounded wings, a bit reminiscent of - but not at all closely related to- the relict in the Sabana Rubia ofthe Sierra de Perija, Dangond dangondi, which we named - lock, stock and barrel' - after the Dangond family who had looked after us so well in northern Colombia (how else - and how better - could a poor teacher repay his gratitude to a rich land­ owning family?). These species have presumably been stuck for so long in their paramo islands that they have e\'olved to become separate genera. Another peculiarity of the paramo browns is the way they demonstrate 'con­ vergent evolution'. In the Cordillera de Merida, R. empetrus and the third paramo species, Altopedaliodes albonotata, both rest on lichen-covered rocks and stones with their wings shut; and, although unrelated, they have very similar underside patterns, right down to the little V- and V-shaped whitish flecks on their hind­ wings. No doubt their identical need for camouflage has resulted in a common pattern, even though their ancestors may have looked thoroughly dissimilar. More striking cases ofconvergence exist in the Santa Marta range (two species of white butterflies from different genera - one ofthem a fascinating relict, Relzquia santamarta - having a very similar pattern ofdarkened veins on the underside of the hindwings) and in the Serrania de Valledupar (the resident white Lymanopoda, L. paramera, looking like an over-sized version ofa little white 'skipper' butterfly, both of which fly together and share brown stripes on the underside that mimic the thin leaves ofthe tussocky grass and bamboos on which they rest). In the mountains of Sama' Marta, Valledupar and Merida, we made some important discoveries about the geographical distributions of the 'pronophiline' brown species. Firstly, the more topographically isolated a mountain range is from the main Andes, the fewer species it has: the Sierra Nevada de Santa Marta, standing alone in the extreme north, has 22 browns; the Serrania de Valledupar, far from the Colombian Eastern C::ordillera but connected to it by land nowhere lower than 1500m, has at least 35; and the Merida range, closer but separated by lower land, has 36. Secondly, although each ofthe ranges is replete with endemic species, many of them are closely related to one another. Thus the Valledupar white Lymanopoda is not nevada, but the paler, more pointed-winged new species, parameraj Pedaliodes tyrrheus in Santa Marta is replaced at similar altitudes by P. tyrrheoides in the Valledupar range and by P. polla in Merida. Darwin would have been delighted if he had known about these examples of geographical speciation! Our third discovery was the most bewildering: as we walked up the mountain­ sides, in the cloud-forests of the Valledupar and Merida ranges, we would find one species suddenly vanishing, to be immediately replaced by a close relative that was obviously distinguishable from it. In some cases, another such transition would occur a few hundred metres higher up, resulting in a series ofthree species in mutually-exclusive bands ofaltitude, spanning the cloud-forest habitat between them. Later on, in the main Colombian Andes, Lymanopoda furnished us with the only examples ofseries ofjour such replacing species, the uppermost one living above the tree-line. Only at the tree-line is there any sign of a change in the vegetation to go with the change in brown butterflies: someone with much more patience than us needs to sit down and study the butterflies in the field, to work 94 TH E .-\LPE\EJ0l.!RNAL out what keeps the species separate; I suspect that the cause might be some aspect of their behaviour towards each other, rather than differences in their habitats, since the climbing bamboos (Chusquea species) on which their lan'ae feed seem to be very wide'spread up the mountainsides. So, the pronophiline browns exhibit a network of allied species, separately inhabiting not only the different mountain 'islands' but also adjacent bands of altitude in the same mountains. How this distribution pattern arose was the subject ofan article ofmine published in 1977; and ofa paper read in October 1983 at a symposium in Peru: the theory proposes a repeated cycle of spread between mountain ranges (in the glacial periods ofthe Pleistocene lee Ages in the past two million years) and divergent evolution in isolation (in the interglacial periods). Thus the adjacent species up the altitude gradients, like the three new species of Pedaliodes in the Serranfa de Valledupar - zuleta (2300-2750m), cesarense (2500-3050m) and vallenata (2950-3200m) - evolved in different mountain areas from a common, more widespread ancestor and, coming together at the next opportunity for spread, they avoided competition by 'splitting up' the available territory into separate zones ofaltitude. The objectives of my last three trips of 1977, 1979 and 1982 (the last two without George Bernard and the last one on my own) were to explore the high-altitude forests and paramo of the three main Colombian Cordilleras. The Eastern and Central Cordilleras are fantastically rich, the former having 87 species of browns and the latter 78 (hardly any of which, by the way, live below 2000m). No burros here, no Amerindians, but friendly peasants, much more rain and endless days immersed in cloud. Now that I knew something about the ways the pronophilines distribute themselves, my task was to fill in gaps in the network. At 2700 m in the Central Cordillera, I thought back to the fauna at the same altitude in the Eastern Cordillera and asked myself, 'Is there a blue Lymanopoda like samius over here? There is no record of such a thing, but there must be some replacement for it'. Sure enough, the tell-tale skipping flight over bamboo clumps signalled a Lymanopoda, and it is brown and white: L. pieridina, known previously from only one specimen erroneously reported to have been caught in the Sierra Nevada de Santa Marta. When I caught a very strange-looking Pedaliodes at 3200m in the Eastern Cordillera, I racked my brains trying to recognise any close relatives; only by examining its genitalia did I see its similarity to a new and as yet undescribed species which [ had discovered at the same elevation in the central chain - a similarity still evident even though the eastern species lacks all the orange markings ofits relative. My mall1 amhition came to be to find all the top-altitude white Lymanopoda species that there are in Colombia, and that meant more than one per mountain chain: the Eastern Cordillera is so broad in the region of Bogota that the high-altitude species on its eastern and western flanks are mutually isolated and some have evolved independently; the same is true in the Tolima (northerly) and Huila (southerly) regions of the Central Cordillera. And in both the cordilleras, not only are the paramo Lymanopoda species white, but the uppermost forest ones are too. One of the latter, L. lactea, is one of those infuriating butterflies that is represented at the Natural History M useum by one solitary specimen, caught in NORTHERN ANDEAN BUTTERFLIES 95 the last century and labelled 'Colombia'. Where in Colombia?! At what altitude? My rediscovery of lac/ea, during a 30-second spell of sunshine at 3100m, just below the tree-line east ofBogota, was the high spot of my 1982 trip and it was worth the long wait during two days ofnon-stop drizzle which preceded it. But yULl COuld go on for ever, battling against the 'law ofdiminishing returns', never satisfied that you have really found them all. Every time you run out offood and leave a good site, some butterfly you have never seen before flaps its wings to 'say goodbye'. Even if it ends up in the net, who is to say there is not another species watching unseen in a forest tree? The species ofPenrosada provide another problem: they are unusual in being distinctly 'seasonal', appearing only at certain, unpredictable times of the year: so we only discovered the two new, endemic Penrosada species in the Serranfa de Valledupar on our fourth visit there, in August 1977. And there still remains the white Lymanopoda mirabilis, named from two improperly-labelled males in 1897 and never seen since. Ifonly the sun had shone that day north-west above Bogota when I patiently quartered the paramo in driving cloud, knowing it to be my last day in Colombia in 1982 and convinced that it was the only place left where mirabilis could be!

I M.J. Adams & G. 1. Bernard (1977) Pronophiline butterflies (Satyridae) ofthe Sierra Nevada de Santa Marta, Colombia. Systematic Entomology, vol.2, 263. 2 M.J. Adams & G. I. Bernard (1979) Pronophiline butterflies (Satyridae) ofthe Serranfa de Valledupar, Colombia-Venezuela border. Systematic Entomology, voI.4,95. 3 M. J. Adams & G. I. Bernard (1981) Pronophiline butterflies (Satyridae) ofthe Cordillera de Merida, Venezuela. Zoological Journal of the Linnean Society, vol.71,343. + M. J. Adams (1977) Trapped in a Colombian Sierra. Geographical Magazine, voI.49,250.

~ --- ...... 96

Sierra Santa Serran!a de Valledupar Caribbean (Sierra de.Perij&) Sea

VENEZUELA

Western Cordillera~~----

Pacific Ocean

BRAZIL

ECUADOR

PERU • = land above 2000m